ABSTRACT
INTRODUCTION: Rivaroxaban concentrations were measured in 127 inpatient samples using an HPLC-MS/MS assay. METHODS: We compared this measurement with a calibrated anti-Xa assay and performed PT, aPTT and dilute PT tests to assess the value of clot-based assays in clinical decision-making. RESULTS: The correlation between the anti-Xa assay and the HPLC-MS/MS at therapeutic concentrations was strong (R2 = 0.98). The PT, RecombiPlasTin 2G, and aPTT, Actin FS, showed a linear dose-response but poor correlation (R2 = 0.32 and 0.44, respectively) and at dilutions of 1 in 150 to 1 in 750 the dilute PT assay also showed poor correlation with rivaroxaban concentrations measured by specific assays. A normal PT or aPTT alone did not identify a likely safe rivaroxaban concentration to allow surgery or invasive procedures, but the combination of normal PT and aPTT identified a group of patients with rivaroxaban levels less than 90 ng/mL. Combined normal PT and aPTT had specificity and sensitivity of 0.97 (95% CI 0.92-0.99) and 0.37 (95% CI 0.1-0.74) for a rivaroxaban concentration < 32 ng/mL. CONCLUSIONS: The PT and aPTT show poor correlation with rivaroxaban levels measured by calibrated anti-Xa and HPLC-MS/MS assays. A normal combined PT and APTT identified low rivaroxaban levels with high specificity but lacked sensitivity. The dPT assay at several dilutions could not be used to quantify rivaroxaban in clinical samples. The utility of these PT, aPTT and dilute PT assays in a clinical setting is very limited, and results generated must be interpreted with caution.